John S. Jacob, Kirana Pandian, Ricardo Lopez, and Heather Biggs

Home , Jacob (Lost), Palustrine wetland, Pothole (landform)

ERPT-002 TAMU-SG-14-303 5/14

Houston-Area Freshwater Wetland Loss, 1992–2010

John S. Jacob Director Texas Coastal Watershed Program

Kirana Pandian Ricardo Lopez Heather Biggs Former Extension Associates Texas Coastal Watershed Program

The Texas A&M University System Acknowledgments

This publication is supported in part by an institutional grant to the Texas Sea Grant College Program (NA10OAR4170099) from the National Sea Grant Office of the National Oceanic and Atmospheric Administration, U.S. Department of Com- merce. The Galveston Bay Estuary Program funded phase one of this work. The SSPEED Center at Rice University funded a portion of the current project. Steven Pennings, Anna Armitage, and Andrew Sipocz provided very helpful reviews. Rebecca Davanon provide cartographic assistance for the final publication. Unless otherwise noted, all photographs were taken by John S. Jacob. Summary

Since 1992, the Greater Houston Metropolitan Area has lost at least 5.5 percent of its natural freshwater wetlands. Although a 5 percent loss in 20 years is unsustainable by any accounting, some areas experienced loss at rates that are catastrophic. For example, Harris County lost almost 30 percent of its freshwater wetlands, including most prominently the iconic prairie pothole-pimple mound complexes. The eight-county Houston area is one of the fastest growing in the country. Metropolitan growth means development, which means the loss of wetlands. When wetlands are lost, with them go important ecological services for both nature and people. Until now we have not had a clear picture of just how many acres of wetlands we might be losing. We report here our attempt to quantify freshwater, non-tidal wetland loss from 1992 to 2010 in the greater Hous- ton region. We developed a simple method for quickly estimating natural freshwater wetland loss. The method, a conservative approach that provides a minimal loss estimate that policy makers can rely on, compares National Wetland Inventory (NWI) maps developed in 1992–93 to more recent digital aerial photography. The study found that most of the losses occurred in rapidly growing Harris, Montgomery, Brazoria, and Fort Bend Counties. The greatest loss occurred in Harris County: 15,855 acres (29 percent) from 1992 to 2010. This acreage loss is more than double that of the other seven counties combined. Losses in other rapidly growing counties, however, are on a track to equal that of Harris County. Overall, the study reveals that development is seriously affecting the freshwater wetland resources in the Greater Houston Metro Area. These freshwater wetlands are the headwaters for virtually all of the water bodies feeding into Galveston Bay. Continued loss at the rates documented here will very likely have grave implications for the long-term health of the Galveston Bay System because it is losing the principal means of cleaning the polluted runoff that enters the bay.

A three-page synopsis on this study, More Flooding, Fewer Fish: Freshwater Wetland Loss in the Houston Area, 1992–2010, is available from the Texas A&M AgriLife Service Bookstore at https://www.agrilifebookstore.org/.

Contents

Introduction ...... 1 Study area ...... 1 Types of freshwater wetlands in the study area...... 2 Methodology...... 3 The tradeoff: sampling versus complete inventory...... 3 Cowardin classification...... 6 Artificial and farmed wetlands...... 7 Logging...... 8 Results ...... 8 Wetland loss in the context of recent Supreme Court rulings on wetlands...... 11 SWANCC ...... 11 Rapanos ...... 11 Implications ...... 12 Loss of services ...... 12 What can be done?...... 13 References...... 14 Appendix A: Wetland loss by county...... 15 Appendix B: Graphs ...... 17 Appendix C: Atlas of wetland loss...... 21 Appendix D: Cowardin Classification ...... 34 Appendix E: Freshwater wetland loss geo-spatial data processing ...... 41 Figures

Figure 1: A coastal prairie pothole wetland in southeast Harris County...... 1 Figure 2: Project study area: the Greater Houston Metropolitan Area...... 1 Figure 3: Barrier island interior wetlands...... 2 Figure 4: Coastal flatwoods wetlands ...... 2 Figure 5: Prairie pothole with encroaching development ...... 3 Figure 6: Riverine forested wetlands, Houston...... 3 Figure 7: NWI wetland polygon converted into Residential classification ...... 4 Figure 8: NWI wetland polygons converted into C/I category...... 5 Figure 9: NWI wetland polygon converted into Fill classification...... 5 Figure 10: NWI wetland polygon converted into Water features...... 5 Figure 11: NWI wetland polygons converted into Mining category...... 6 Figure 12: Farmed wetlands indicated in northwest Harris County...... 7 Figure 13: Farmed wetland delineation Chambers and Liberty Counties...... 7 Figure 14: Farmed wetlands with potholes...... 7 Figure 15: (a) Forested wetland in a deforested state, and (b) forested wetland after 6 years in a regrown state ...... 8 Figure 16: Graph of wetland loss by county...... 9 Figure 17: Total natural freshwater wetland loss in Greater Houston Metro Area...... 9 Figure 18: Harris County wetland loss in acreage and percentage by each period ...... 9 Figure 19: The White Oak Bayou watershed lost 71 percent of its wetlands (shown in red) from 1992 to 2010...... 9 Figure 20: The Cypress Creek area lost 29 percent of its wetlands (shown in red) from 1992 to 2010 ...... 9 Figure 21: The Spring Creek area lost 11 percent of its wetlands (shown in red) from 1992 to 2010 ...... 10 Figure 22: Harris County wetland loss under FEMA 100-year floodplain...... 11 Figure 23: A prairie pothole marsh in Harris County ...... 13 Figure 24: A freshwater marsh ...... 14

Tables

Table 1: Palustrine wetlands classes ...... 6 Table 2: Wetland loss by county...... 8 Table 3: Wetland loss by type of destruction...... 10 Table 4: Total wetland loss by wetland type, Greater Houston Metro Area ...... 10 Table 5: Wetland loss with respect to the FEMA 100-year floodplain...... 12 Introduction

The population of the Greater Houston Metro- While somewhat imprecise because of the limita- politan Area is skyrocketing, particularly in Hous- tions of the source data, this method is more than ton and in the counties nearest to it1. Between 1990 sufficiently reliable as to be actionable. and 2010, it added more than 2 million2 inhabi- tants, and 3.5 million to 4 million more people are Study area forecast for the next 20 to 30 years3. This growth has obvious impacts on wetland resources. The study area consists of the eight counties that make up the Greater Houston Metropolitan Area: The U.S. Environmental Protection Agency Brazoria, Chambers, Fort Bend, Galveston, Harris, (EPA) defines wetlands as “areas where water covers Waller, Montgomery, and Liberty Counties (Fig. 2). the soil, or is present either at or near the surface of the soil all year or for varying periods of time during the year, including during the growing season.”4 Wetlands are “transition zones”5 that are neither land nor water but have characteristics of both as well as properties of neither (Fig. 1). A wetland area need not be wet all year long—as few as 2 or 3 consecutive weeks of wetness can qual- ify an area as a wetland. However, most wetlands in the Houston area are wet for substantially longer periods. Wetlands provide many functions, including detaining stormwater, controlling erosion, stor- ing and cleansing water, and providing places for recreation for people and habitat for wildlife. When Figure 1: A coastal prairie pothole wetland in southeast Harris wetlands are destroyed, these services are also lost. County (Source: Cliff Meinhardt) It is possible, however, for urban growth and wetlands to coexist. Among the strategies for protecting wetlands is to change the way that cities and their outlying regions are developed. But to take action, residents, businesses, and policy makers need reliable information on the extent of the problem. How many wetlands are actually being lost? Is this decrease significant? Where are the remaining wetlands that could be protected? The Texas Coastal Watershed Program, a joint program of Texas A&M AgriLife Extension and Texas Sea Grant College Program devised a method to gather semi-quantitative data about wetland loss in the eight-county region surrounding Houston.

1 H-GAC-2035, August 2006, 2035 Regional Growth Forecast 2 US Census Bureau, March 2013, State and County Quick Facts 3 H-GAC-2035, Aug 2006, 2035 Regional Growth Forecast 4 U.S. Environmental Protection Agency. 1995. America’s Wetlands: Our Vital Link Between Land and Water. EPA843-K-95-001 5 Moulton, D.W, and J. S. Jacob. 2000, Texas Coastal Wetlands Figure 2: Project study area: the Greater Houston Metropolitan Guidebook Area

1 Types of freshwater wetlands big enough to maintain this kind of flooding. In in the study area most bottomland hardwood forests, flooding lasts for a few weeks to several months. Wetlands differ in hydrology, which is the Swamps, on the other hand, stay flooded much movement, distribution, and quality of water. Four longer; they may dry out only occasionally. Swamp types of natural freshwater wetlands occur in the forests occur in sloughs and other depressed areas study area: barrier island interior wetlands, coastal of floodplains and occasionally in low floodplain flatwoods wetlands, prairie pothole wetlands, and terraces, where the dominant source of water may riverine forested wetlands. For a map of where the be runoff from rainfall. different types of wetlands occur in this region, see the Texas Coastal Wetlands Guidebook by D. W. Moulton and J. S. Jacob. Barrier island interior wetlands (Fig. 3) are fed by a combination of groundwater and runoff from nearby dunes. The water seeps easily through the sandy dunes and generally surfaces in the swales between the dunes. Although water rarely ponds in many of these swales, the groundwater remains just under the surface for extended periods. Coastal flatwoods wetlands receive water primarily from local rainfall that runs off or seeps into the soil very slowly. They are typically wet in winter and early spring. The soil is saturated, and shallow water stands in many places. Figure 3: Barrier island interior wetlands (Source: Earl Nottingham, Texas Parks Prairie pothole wetlands (freshwater and Wildlife Department) marshes) receive water by direct rainfall and by runoff from surrounding flats. Some prairie potholes may have groundwater as a water source, particularly in sandy areas such as barrier islands. Water movement and storage in pothole complexes are very diverse, with distinct differences occurring within just a few feet. Deeper potholes can remain saturated for more than 6 months of the year; nearby pimple mounds may be nearly semi-arid for most of the year. This hydrologic com- plexity engenders high habitat and biological diversity. Riverine forested wetlands receive water primarily from overbank river flooding. Flooding occurs in most years and persists for at least several weeks at a time. Only larger rivers have watersheds Figure 4: Coastal flatwoods wetlands(Source: Andrew Sipocz, TPWD)

2 Figure 5: Prairie pothole with encroaching development Figure 6: Riverine forested wetlands, Houston

Methodology

The tradeoff: sampling versus only capture wetlands during their normal period complete inventory of maximum inundation, but they also plainly dis- tinguish between the typically deciduous nature of To estimate the amount of wetland loss, we estab- wetland vegetation and the often evergreen nature lished a baseline on which to compare subsequent of upland vegetation in the Houston region. trends. The National Wetland Inventory (NWI), We did not have access to the original photography conducted periodically by the U.S. Fish and Wild- used by the NWI project. No ground-truthing (con- life Service (wetlands.fws.gov), produces maps that firming information from photographs by gathering could serve as a potential baseline in many parts of data in the field) took place as part of this audit. the country. On average, this method found 50 to 75 percent To determine whether the NWI maps were more wetlands than what the NWI mapped on the accurate for the Houston area, we compared them same quarter quadrangles. No areas that the NWI to maps we created using 1995 color infrared (CIR) mapped as wetlands had anything other than a wet- aerial photographs supplied by the Texas Natural land signature on the aerial photography. This find- Resource Information System. We used a random ing confirmed our field experience of many years, number generator to select three USGS quar- and that of others, that the NWI underestimates the 6 ter-quadrangles from throughout the study area , wetlands on the ground, and that very few if any and an experienced wetland mapper (the senior NWI maps indicate wetlands where wetland signa- author) interpreted the photographs of these quar- tures were lacking on the aerial photos. ter-quadrangles. The digital aerial photography had much higher Because the 1995 photographs were taken in win- resolution than that available to the NWI staff in ter under wet to normal conditions, the wetlands 1992. Also, the later photographs were taken in a were relatively distinct. Winter photographs not wetter year, when more wetlands would be distin- 6 This assessment did not examine areas of “farmed wetlands” guishable. Even under the best of circumstances, (PEMf). This is a problem class for this area as discussed below. we would not expect the NWI to have matched the

3 accuracy of our later efforts. The important point is 10.1. For a detailed description of the GIS method- that the NWI maps identify precisely the wetlands ology, see Appendix B. that were observable using the photography avail- We were extremely cautious in identifying devel- able at the time. There were very few false positives opment in the aerial photos. The obvious cases of in the NWI maps7. strip malls, residential developments, and the like The NWI survey for the Greater Houston Metro posed no interpretive challenge. More difficult were Area thus constitutes an acceptable baseline for vegetation removal and/or excavation/fill without assessing wetland loss. The NWI grossly underesti- further development. Only in cases of obvious wet- mates the true amount of wetlands on the ground, land destruction did we classify a wetland as filled. but there is a very high chance (we estimate more Our assessment of wetland loss is thus a very than 95 percent) that the NWI wetlands mapped conservative assessment. The loss estimates are indeed wetlands. reported here are not maximal, but minimal. Whether this sample is statistically representative We subdivided development into the following in every way we cannot say, but we have no basis to categories: assume that it is not representative. Without ques- ■ Residential: Generally homes and other resi- tion, it constitutes a minimal estimate of the total dences; some light commercial and roads amount of wetlands at the time. An analysis of loss ■ Commercial/industrial: Malls, strip malls, of NWI wetlands thus gives policy makers a reliable industrial and commercial facilities bracket around the true value of wetland loss. ■ Fill: Undefined fill; obvious removal of vegeta- To identify changes in land use, we simply over- tion; excavation laid the digitized NWI lines from the latest year ■ Water: Wetlands replaced by an open water available at the time of this study (generally 1992 feature, such as a pond or lake or 1993)8 onto the latest digital aerial photography available (2010) and determined the amount of ■ Mining: Mining activity, generally gravel wetlands that had been lost to development. This Figures 7 through 11 depict the editing process procedure was accomplished using ArcGIS Desktop and examples of each development type.

Figure 7: Left: 1995 aerial photo with NWI polygon (TNRIS color photograph). Right: 2010 aerial photo of the same area with the same superimposed NWI polygon, with obvious new development. (TNRIS CIR photography)

7 Except for the special case of the palustrine farmed wetland class discussed in “Artificial and farmd wetland” on page 7. 8 The NWI has recently been updated for Galveston County and the rest of the circum-Galveston Bay quadrangles. This report uses the 1992–93 lines as the baseline for wetland loss.

4 Figure 8: As above, with industrial development

Figure 9: As above, with fill activity but no additional development beyond initial clearing and excavation. Note the small area of wetland not destroyed in the lower right of the photo on the right. This remaining area is obviously negatively impacted, but for the purposes of this report it was not considered a wetland loss.

Figure 10: As above, with land use changed to a water feature, which does not function as a wetland, and has nowhere near the ecological value of a wetland

5 Figure 11: As above, mining activity. Note that some 1993 wetlands had already been lost to mining in 1996.

Cowardin classification Subclasses are based on hydrology, water chemistry, and the nature and persistence of vegetation. The Cowardin wetland classification is com- Subclasses are indicated by a series of letters or monly used throughout the United States and is the numbers after the class level. For example, PFO2T system used by the NWI. It is a hierarchical system refers to a palustrine forested needle-leaved decid- based primarily on hydrology and vegetation, and uous tidally influenced wetland (such as a cypress secondarily on the nature of the bottom or sub- swamp near the mouth of a river). strate. The entire Cowardin classification is reproduced This report focuses only on the palustrine in Appendix D. (non-tidal inland freshwater wetlands) wetland system, the highest unit in the Cowardin Table 1: Palustrine wetland classes scheme. Class Class ID Name Description Riverine wetlands are Palustrine PAB Aquatic bottom Submergent vegetation limited to river channels; they aquatic bed comprise a tiny percentage of Palustrine PEM Emergent Herbaceous vegetation the study area. The lacustrine, emergent or lake, system is also of rela- Palustrine PFO Forested Wooded areas tively small percentage of the forested study area. Most freshwater wetlands in the study area are Palustrine scrub PSS Scrub-shrub Usually secondary growth of the palustrine system (P). shrub (such as Chinese tallow tree or shrubby vegetation) Its class taxa, based mainly on vegetation, are given in Table Palustrine PUB Unconsolidated Very little/no vegetation, 1. The only significant palus- unconsolidated bottom flooded wetlands subject to trine wetlands in the study bottom wave and current action area are palustrine emergent, Palustrine PUS Unconsolidated Sparse vegetation, flooded palustrine forested, and palus- unconsolidated shore shoreline wetlands, sandy or trine scrub shrub (Table 1). shore muddy

6 Artificial and farmed wetlands Figure 13 shows some farmed wetland areas in Chambers and Liberty Counties. The NWI maps show artificial as well as natural The farmed wetland areas form very large PEMf wetlands. This study focuses only on natural wet- delineations in these areas. They were likely flooded lands. We excluded diked, excavated, spoil, and arti- at the time the NWI mapping occurred. When the ficial substrate wetlands. The excluded wetlands are 1992 NWI team mapped these areas, they might for the most part the result of human construction. have been either rice fields or fallow-year inunda- Farmed wetlands are of special interest. The NWI tions for attracting waterfowl. They are clearly much used the farmed prairie wetland category (PEMf larger than the natural prairie potholes in the area. in the NWI; see Appendix D) to map both natural These large areas should have been categorized wetlands that were farmed as well as large areas that as diked/impounded instead of farmed because the were diked off for rice or for temporary waterfowl entire area is clearly not a permanent wetland. habitat. Figure 12 shows the location of farmed wetlands (PEMf) in northwest Harris County, and The PEMf taxon covers large areas (about 173,787 acres, or 51 percent of the total PEM coverage of the study area; see the insets of Figs. 12 and 13). Clearly, bona fide wetlands are in each large PEMf polygon, but our project did not measure them. Figure 14, taken from Google Earth, shows a large farmed wetland delineation on a 2010 aerial photograph with relatively few pothole wetlands, as compared to the larger delineation. The actual wetlands in that polygon appear to be less than 20 percent of the area. Because this farmed wetlands class vastly overmaps bona-fide wetlands, we simply removed this category from our analysis9. Our analysis did include farmed wetlands that were marked with the special modifier “d” (such as Figure 12: NWI delineations of farmed wetlands (PEMf) in PEMd) for drained wetlands. Most of these drains northwest Harris County were temporary, such as for draining rice fields. In these cases, the wetland depressions remain intact, as does their long-term hydrology.

Figure 13: NWI delineations of farmed wetlands (PEMf) cover- ing up half of Chambers County and parts of Liberty County Figure 14: The yellow polygon is a PEMf delineation from the 1992 NWI data. This PEMf polygon is clearly overextended. 9 The issue with PEMf wetlands should not be confused with the Only the small dark areas (prairie potholes) are bona fide general underestimation of natural wetlands found in the NWI. wetlands. 2010 Google Earth photo.

7 Logging We did not categorize any of the areas showing extensive tree harvest (Fig.15a) as wetland losses, Logging occurs in some of the forested wetlands unless they were converted to development or were of the study area, mostly in Liberty and Cham- very obviously filled. bers Counties and parts of Montgomery County. However, unless fill occurs, there is no reason to assume that logging destroys wetlands. For example, although some soil was disturbed as the result of logging that left the denuded or treeless area in Figure 15(a), the regrowth seen in the same spot in Figure 15(b) strongly suggests the absence of fill activity or major soil disturbance.

Figure 15(b): The same 1993 NWI PFO polygon in Fig. 15(a) superimposed on a 2012 aerial image from Google Earth. Trees are clearly growing, and the linear rows of planted trees are evident in the photo. A classification by means of digital algo- rithms, such as the one that the NOAA Coastal Change Analysis Program (C-CAP) uses, would have labeled the 2006 image a land use change, which clearly it is not, at least not on a 10–20 year time scale of tree growth. The 2006 image merely captures the area soon after a timber harvest. Thus it would be improper to count this polygon as a filled or lost wetland. That this wetland has been manipulated by harvest of the trees there is little Figure 15(a): 1993 NWI PFO (Palustrine forested) polygon doubt. But without intensive ground truthing, we cannot label superimposed on a 2006 Image from Google Earth. The area the tree harvest as a wetland loss and still be consistent with has clearly been logged. But is this a wetland loss? the conservative approach we have taken in this project.

Results

The study found that by 2010, the Greater Hous- Table 2: Wetland loss by county ton Metro Area lost about 5.5 percent, or 24,600 County Total Loss in Loss as of the 447,949 acres of natural freshwater wetlands acres acres percentage of mapped by NWI in 1992–93 (Table 2, Fig. 16). county wetlands Most of the losses occurred in rapidly growing Brazoria 88,838 855 0.9% Harris, Montgomery, Brazoria, and Fort Bend Chambers 58,041 255 0.4% Counties. The greatest losses by far were in Harris Fort Bend 24,002 1,442 6% County (Fig. 17), where almost 30 percent of the freshwater wetlands existing in 1992 have disap- Galveston 14,316 1,066 7.4% peared. Harris County lost 15,855 acres of wetlands Harris 54,479 15,853 29.1% (Fig. 18), almost twice the amount of the other Liberty 165,987 1,951 1.2% seven counties combined (Fig. 16). Because these Montgomery 32,498 3,093 9.5% acreage figures are a minimal estimate, the actual Waller 9,788 83 0.8% number of acres lost could be well over 45,000 (see discussion above on NWI accuracy under “Meth- Total odology”). (all counties) 447,949 24,600 5.5%

8 The area of greatest loss in the study was western Harris County wetland loss Harris County, where: ■ The White Oak Bayou watershed lost 71 per- 30.0% 5.6% cent of the wetlands that were present in 1992 3,027 (Fig. 19) 25.0% 20.0% 10.4% ■ The Cypress Creek watershed lost 29 percent 5,659 Year 2006–2010 15.0% (Fig. 20) Year 2002–2006 10.0% 13.2% ■ The Spring Creek watershed, 11 percent (Fig. 21) 7,168 Year 1992–2002 5.0% Loss in acres Because White Oak Bayou is closer to the center of 0.0% Houston, it would be expected to have more development. However, if current trends continue, wetland All wetlands loss in the Cypress Creek and Spring Creek water- Figure 18: Harris County wetland loss in acreage and percent- ! sheds will soon equal that of White Oak Bayou, with age by each period the potential for consequent increases in flooding and reduction of water quality in the creeks.

Wetlands Loss by County 15,854 30.0% 25.0% 20.0% 15.0% Loss in acres 1,06666 10.0% 3,09309093 1234 & 1,4421 442442 5.0% From 1992–2010 1,9511 Total acres lost: 24,600 0.0% 855 83 254

Figure 16: Graph of wetland loss by county, with percentage loss within each county Figure 19: The White Oak Bayou watershed, which lost 71 percent of its wetlands (shown in red) from 1992 to 2010

Figure 17: Total natural freshwater wetland loss in the Greater Houston Metro Area. Green areas are undeveloped wetlands. Figure 20: The Cypress Creek watershed, which lost 29 percent Red areas are developed or filled wetlands as of 2010. of its wetlands (shown in red) from 1992 to 2010

9 We performed some intermediate assessments for Harris County. Figure 17 shows loss of wetlands in Harris County over three separate periods in the past 18 years. On a per-year basis, the period from 2002 to 2006 had the highest rate of loss, 1,400 acres, or about 2.6 percent per year. The other two periods had similar rates of loss, 600 to 700 acres, or about 1.3 percent per year. The 2006 to 2010 period included the beginning of the Great Recession, such that the number of acres lost in the first 2 years of that period was probably similar to that of the 2002–06 period. The next highest wetland losses were 9.5 percent in Montgomery County and 7.4 percent in Gal- veston County. As the economy recovers from the Figure 21: The Spring Creek watershed, which lost 11 percent of Great Recession of 2008, losses in the high-growth its wetlands (shown in red) from 1992 to 2010 Houston metro counties will probably equal those of Harris County. Table 3: Wetland loss by type of destruction Of the total losses for the entire study area, 64 Wetland loss type NWI lost (acres) NWI lost (%) percent could be attributed to completed develop- Residential (R) 12,805 52% ment projects (Table 3), and 27 percent were wet- Commercial/ industrial (I) 2,989 12% lands that had been filled and destroyed but had not been developed. Figure 21 shows a map of the Filled (F) 6,622 27% relative loss of the total natural freshwater wetlands Water (W) 828 3.4% across the entire study area, and Table 4 shows the Mined (M) 1,354 5.5% loss by wetland type in the study area. Total loss numbers for each county by Cowardin class are in Appendix A. Table 4: Total NWI wetland loss by wetland type, Greater Houston Metro Detailed maps of wetland loss are in the Area atlas in Appendix C. System- Description Total Acres % wetland This study documents the severity of Class acres lost loss the impact of development on freshwater PAB Palustrine aquatic bed 1,857 29 1.5% wetland resources in the Greater Houston PEM Palustrine emergent 141,678 6,168 4.3% Metro Area. These freshwater wetlands are the “headwaters” for virtually all of PFO Palustrine forested 276,695 15,562 5.6% the water bodies feeding into Galveston PSS Palustrine scrub-shrub 23,239 2,751 11.8% Bay. Continued loss at these rates will PUB Palustrine unconsolidated 4,310 75 1.7% have grave implications for the long-term bottom health of the Galveston Bay system. PUS Palustrine unconsolidated 170 15 8.9% shore Total 447,949 24,600 5.5%

10 Wetland loss in the context of recent Supreme Court rulings on wetlands

Since 2001, the U.S. Army Corps of Engineers Section 404 protection (those outside of the 100- (USACE) has deemed the vast majority of wetlands year floodplain). Most of the palustrine emergent documented as lost in this study as outside of its wetlands (PEM, marsh or “prairie-pothole wet- jurisdiction. The Galveston District of the USACE lands”) are outside of the 100-year floodplains and, currently considers almost all palustrine wetlands therefore, for the most part outside of Clean Water subject to development in this area to be “isolated” Act jurisdiction, under current USACE Galveston from the “waters of the U.S.” District interpretations. The study period for this project straddles two Eighty-three percent of PEM wetlands lost in the major U.S. Supreme Court rulings dealing with study area occur outside of the 100 year floodplain how wetlands are regulated. Both of these decisions (Table 5). center on a legal doctrine known as the “significant nexus.” The fundamental question is how con- Rapanos nected a wetland is to “a water of the U.S.,” a water body within the regulatory authority of the federal In 2006, the Supreme Court in Rapanos v. United government under the Clean Water Act. If a wet- States, 547 U.S. 715 (2006) further clarified the sig- land is clearly connected to a water of the United nificant nexus rule. If a particular class of wetlands States, it is very likely to be regulated. could be demonstrated to have a significant nexus to waters of the United States, then that class of SWANCC wetlands could be considered jurisdictional without having to document the connection for every case. In 2001, the U.S. Supreme Court ruled in Solid In response to this ruling, the hydrology of Waste Agency of Northern Cook County (SWANCC) coastal plain prairie potholes on the Upper Gulf v. the Army Corps of Engineers, 531 U.S. 159 (2001), that the Migratory Bird Rule was not a sufficient reason for reg- ulating an otherwise isolated wetland. After this decision, the Galveston District USACE defined hydrologically isolated wetlands narrowly, which rendered almost all wetlands outside of the FEMA 100-year floodplain in this district exempt from regulatory jurisdiction. The exceptions were the very few wetlands outside the floodplain with a “bed and banks” connection—a virtual river bed—to a floodplain or a water of the United States. Figure 22 shows the distribution of palustrine wetlands (prairie and forested potholes) and FEMA 100-year floodplains in Harris County in the area of greatest wetland loss. The map Figure 22: Harris County wetland loss under FEMA 100-year floodplain. gives a sense of the amount of wet- Red areas are NWI wetlands that have been lost to development. The most lands no longer under Clean Water Act recent FEMA 100-year floodplain overlay is also shown.

11 Coast of Texas was investigated. The studies Table 5: Wetland loss with respect to the FEMA 100-year floodplain revealed that an average of about 15 to 20 County Total Total Acres lost % lost percent of the precipitation and runoff that wetland acres in 100-year in 100-year flows into these pothole wetlands flows into acres lost floodplain floodplain waters of the United States (Wilcox et al. Brazoria 88,838 855 209 24% 2011, Forbes et al. 2012). Chambers 58,041 255 31 12% As of this writing, the Galveston District Fort Bend 24,002 1,442 237 16% of the USACE has not modified any policies Galveston 14,316 1,066 232 22% based on these studies. However, a wetland Harris 54,479 15,855 2,165 14% connectivity review recently published by the EPA specifically addresses the so-called Liberty 165,987 1,951 527 27% isolated wetlands that make up the losses Montgomery 32,498 3,093 845 27% documented in this study. The review notes Waller 9,788 83 25 31% the relatively large contribution of water to Subtotal 447,949 24,600 4271 17% surface streams these wetlands supply. The Source: National Flood Hazard Layer (NFHL) dataset from FEMA, 2012, Harris, Waller, science supports exerting federal jurisdic- and Liberty Counties FEMA Q3 1996 A zone floodplain data – Fort Bend, Montgomery, Galveston, Brazoria, tion over these wetlands (USEPA 2013). and Chambers Counties

Implications

Loss of natural freshwater wetlands in the ■ Soaking up water to help replenish groundwa- Greater Houston Metro Area over the 18 years of ter supplies the study period (1992 to 2010) was massive and ■ Reducing erosion rapid. As shown in Table 4, the area lost 4.3 percent ■ Providing habitats for wildlife, including of the most endangered category of wetlands in migratory birds the overall area, the palustrine freshwater marshes ■ Protecting coastal areas and shorelines by (PEM, prairie potholes in the local parlance). weakening the force of storms In Harris County, however, a staggering 36 ■ Providing places for boating, fishing, hiking, percent of its prairie marshes was lost (Appendix hunting, and bird-watching A), accounting for more than 70 percent of the total ■ Offering an intangible sense of beauty and loss of the prairie marshes in the entire study area. place in our culture As the economy improves, even more wetlands will In the Houston area, wetlands are the principal be filled, and few losses will be mitigated. means of cleaning polluted runoff that enters Gal- Wetlands in Harris County are being lost so veston Bay. quickly that little can be done except to try to save The cost of replacing those services would be astro- a few critical last pieces of ecologically significant nomical. For example, the wetland loss in the study land. Counties surrounding Harris County can area is equivalent to at least 12,000 acre-feet, or nearly expect similar losses in the next few years. 4 billion gallons, of stormwater detention. This stormwater detention is in addition to that provided by the Loss of services native soils. At an average cost of $50,00010 per acre- Freshwater wetlands provide critical ecological foot of stormwater detention, this loss corresponds to services to the Galveston Bay System: no less than $600 million. For Harris County, the loss ■ Decreasing flooding is at least 7,000 acre-feet, or $350 million. ■ Cleaning runoff water by filtering out pollution 10 Based on Harris County Flood Control figures courtesy of Carolyn and sediment White for some recent mid-sized detention basins in Harris County.

12 These values are just for stormwater detention. permits include evidence that they have a CWA They do not take into account the water-cleansing permit or that the USACE does not recognize value of the wetlands or any of the other benefits any wetlands on the site. Much of the develop- mentioned above. Counting these services, the loss ment in the region occurs without any inves- is clearly in the billions. tigation of the presence of wetlands, much less Much of what is being lost now is among the with a Clean Water Act wetland permit. We most valuable habitat remaining on the entire suspect that half or even more of all develop- Upper Texas Gulf Coast. Vast acreages of land ment occurs this way. were land-leveled for agriculture during the 20th ■ Cities and other jurisdictions could also century. Some of the best examples of undisturbed require that the mitigation occur in the same prairie pothole-pimple mound complexes are in watershed as the development, thus enabling urban fringe areas yet to be developed and in areas more green space in or near the jurisdiction, where agriculture has not yet penetrated. These are without any additional costs to the citizens. the wetlands now under the greatest threat. ■ Houston is set to gain another 3 million to 4 million people in the next 30 years. New devel- What can be done? opment might cover 700 to 1,000 square miles. If the coastal prairie pothole wetlands, which Wetlands are essential to the long-term ecologi- make up most of the freshwater wetlands in the cal stability of our entire region. They are a natural eight-county region, were fully protected under legacy that we can ill afford to lose. And yet we are the CWA and their development mitigated losing them at an increasing pace as our economy appropriately, 25,000 to 35,000 acres of wetland revives. Now that we know how many wetlands we habitat, the equivalent of an Anahuac National are losing and where we are losing them, we can Wildlife Refuge, could be set aside every year. begin to consider the alternatives to our current ■ way of doing business: Coastal resource managers need to identify the critical habitat that remains and work with res- ■ Because freshwater wetlands are being lost pri- idents and conservationists to develop plans for marily to development, many wetlands would preserving and restoring wetlands. The Texas be preserved if cities and counties simply Coastal Watershed Program, a joint program required that developers applying for building of Texas A&M AgriLife Extension and Texas Sea Grant College Program, in conjunction with the Houston–Gal- veston Area Council, has mapped out the remaining ecologically significant areas of high quality prairies and forests that have an abundance of wetlands (http://arcgis02.h-gac.com/ ecologicalGIS/). ■ Coastal resource managers and other conservationists also need to consider how wetland mitigation could play a much larger role in ecosystem resto- ration and maintenance in the region. ■ Resource managers must work with local citizens to educate them on the implications of wetland loss in our area. Without citizen support, little can be done to preserve critical areas on Figure 23: A prairie pothole marsh in Harris County (Source: Andrew Sipocz) the scale that is needed.

13 ■ Regulators who oversee the CWA permitting or has been obtained for development on a process should carefully consider recent specific property. scientific studies documenting the hydrologic ■ Citizens can also work with county and munic- connection of freshwater wetlands in the ipal governments to require wetland mitigation study region to Galveston Bay and other water with their drainage and detention permitting bodies. Their decisions could greatly affect the for medium and large developments. quality of our coastal waters and ecosystems. The wetlands that beautify our region and sus- ■ Finally, residents can be involved by reporting tain us are an irreplaceable legacy of nature handed unauthorized wetland fill activities directly to down to us by previous generations. If the current the USACE Galveston District office. A citizen loss proceeds at its current increasing pace, we will can ask the Corps whether a permit is needed be the generation responsible for the loss of this most precious resource.

References

Cowardin, L. M., V. Carter, F. C. Golet, and E. T. U.S. Environmental Protection Agency. 2013. Con- LaRoe. 1979. Classification of Wetlands and Deep- nectivity of Streams and Wetlands to Downstream water Habitats of the United States. Biological Waters: A Review and Synthesis of the Scientific Services Program. U.S. Fish and Wildlife Ser- Evidence. EPA/600/R-11/098B. vice. FWS/OBS-79/31. U.S. Government Printing Wilcox, B. P., D. D. Dean, J. S. Jacob, and A. Sipocz. Office. Washington, DC. 2011. “Evidence of Surface Connectivity for Texas Forbes, M. G., J. Back, and R. D. Doyle. 2012. Gulf Coast Depressional Wetlands.” Wetlands, “Nutrient Transformation and Retention by 31(3), 451–458. Coastal Prairie Wetlands, Upper Gulf Coast, Texas.” Wetlands, 32(4), 705–715. Jacob, J. S., and R. Lopez. June 2005. Freshwater, Non-Tidal Wetland Loss, Lower Galveston Bay Watershed 1992–2002. A Rapid Assessment Method Using GIS and Aerial Photography. Contract Report No. 582-3-53336. Galveston Bay Estuary Program. Texas Coastal Watershed Program. Moulton, D. W., and J. S. Jacob. 2000. Texas Coastal Wetlands Guide- book. Texas Sea Grant Publication TAMU-SG-00-605 (r). State and County Quick Facts: U.S. Census Bureau. http://quickfacts.census.gov/ qfd/states/48000.html Accessed May 2013. Figure 24: A freshwater marsh

14 Appendix A: Wetland loss by county

Abbreviations used in the following tables: ■ PAB: Palustrine aquatic bed ■ PEM: Palustrine emergent ■ PFO: Palustrine forested ■ PSS: Palustrine scrub shrub ■ PUB: Palustrine unconsolidated bottom ■ PUS: Palustrine unconsolidated shore

Harris County wetland loss by system-class Wetland Total Year 1992–2002 Year 2002–2006 Year 2006–2010 Year 1992–2010 class acres Acres lost % Acres lost % Acres lost % Acres lost % PAB 78.6 18 22.8% 6.8 8.6% 1.7 2.2% 26.4 33.6% PEM 12,474 2,259.7 18.1% 1,413.5 11.3% 804.7 6.4% 4,477.2 35.9% PFO 37,137.5 4,033 10.9% 3,522 9.5% 1,905.3 5.1% 9,459.1 25.5% PSS 4,309.7 834.1 19.4% 688.4 16% 300.4 7% 1,822.7 42.3% PUB 411 19.4 4.7% 22.8 5.6% 10.5 2.5% 52.7 12.8% PUS 68.4 3.6 5.2% 5.6 8.2% 5.2 7.6% 14.4 21% Total 54,479.3 7,167.7 13.2% 5,659.1 10.4% 3027.7 5.6% 15,852.5 29.1%

Galveston County wetland loss by system-class Wetland Total Year 1992–2002 Year 2002–2010 Year 1992–2010 class acres Acres lost % Acres lost % Acres lost % PAB 6 – – 1 16.5% 1 16.5% PEM 11,123.6 96.5 0.9% 297.4 2.7% 393.9 3.5% PFO 1,867.1 88.2 4.7% 315.1 16.9% 403.3 21.6% PSS 1,187.4 70.6 5.9% 193.5 16.3% 264 22.2% PUB 97.3 2.2 2.3% 0.9 0.9% 3.1 3.2% PUS 34.8 – – 0.5 1.5% 0.5 1.5% Total 14,316.2 257.5 1.8% 808.3 5.6% 1065.8 7.4%

15 Brazoria County wetland loss by system-class Liberty County wetland loss by system-class Wetland Total Year 1992 –2010 Wetland Total Year 1992 –2010 class acres Acres lost Percentage class acres Acres lost Percentage of loss of loss PAB 610.9 – – PAB 622 – – PEM 44,398.4 173.9 0.4% PEM 16,557.3 162.6 1% PFO 40,517.9 570.3 1.4% PFO 13,9485.1 1,673.1 1.2% PSS 2,963.2 110.5 3.7% PSS 8,479.4 113.6 1.3% PUB 343.6 0.7 0.2% PUB 827.4 2.2 0.3% PUS 4 – – PUS 15.4 – – Total 88,838.1 855.3 1.0% Total 165,986.6 1951.5 1.2%

Chambers County wetland loss by system-class Montgomery County wetland loss by system-class Wetland Total Year 1992–2010 Wetland Total Year 1992 –2010 class acres Acres lost Percentage class acres Acres lost Percentage of loss of loss PAB 124.5 – – PAB 114.7 1.4 1.2% PEM 41,466 32.3 0.08% PEM 5,943.4 536.6 9% PFO* 12,852.8 182.7 1.4% PFO 23,410.2 2,285.6 9.8% PSS* 2,378.4 39.3 1.4% PSS 2,597.5 261 10% PUB 1,190.7 0.2 0.02% PUB 425.9 8.7 2.1% PUS 28 – – PUS 6.3 – – Total 58,040.5 254.5 0.4% Total 32,498.1 3,093.2 9.5%

Fort Bend County wetland loss by system-class Waller County wetland loss by system-class Wetland Total Year 1992 –2010 Wetland Total Year 1992 –2010 class acres Acres lost Percentage class acres Acres lost Percentage of loss of loss PAB 278.4 – – PAB 21.5 – – PEM 5,662.4 322.5 5.7% PEM 4,052.4 67.9 1.7% PFO* 16,543.0 980.4 5.9% PFO 4,881.7 6.7 0.1% PSS* 749.5 136.7 18.2% PSS 573.4 3.2 0.6% PUB 762.9 2.2 0.3% PUB 251.5 4.9 2% PUS 6.1 0.4 6.5% PUS 7.3 – – Total 24,002.4 1,442.2 6% Total 9,787.8 82.7 0.8%

16 Appendix B: Graphs

Abbreviations used in the following graphs: ■ PAB: Palustrine aquatic bed ■ PEM: Palustrine emergent ■ PFO: Palustrine forested ■ PSS: Palustrine scrub shrub ■ PUB: Palustrine unconsolidated bottom ■ PUS: Palustrine unconsolidated shore

Wetland loss for each county 15,854 30.0%

25.0%

20.0% Loss in acres 15.0% ! 1,0666 10.0% 3,093933 1,442 From 1992–2010 5.0% Total acres 1,951 lost: 24,600 0.0% 855 83 254

Fort Bend

Brazoria County wetland loss

4.0% 110 3.5% 3.0% Loss in acres 2.5% 2.0% 570 1.5% Loss from 1992–2010 1.0% 174 0.5% 0.0% PEM PFO PSS

17 Chambers County wetland loss

1.80% Loss in acres 39 1.60% 183 1.40% 1.20% 1.00% 0.80% Loss from 1992–2010 0.60% 0.40% 32 0.20% 0.00% PEM PFO PSS

Fort Bend County wetland loss

20.0% 137

15.0% Loss in acres

10.0% 322 980 Loss from 1992–2010 5.0%

0.0% PEM PFO PSS

Galveston County wetland loss

25.0% Loss in acres 20.0% 15.0% 315 10.0% 193 Year 2002–2010 5.0% 297 Year 1992–2002 96 88 0.0% 71 PEM PFO PSS

18 Harris County wetland loss

45.0% 40.0% Loss in acres 35.0% 2,260 805 30.0% 25.0% 1,413 688 Year 2006–2010 20.0% 1,413 Year 2002–2006 15.0% 3,522 10.0% 2,260 Year 1992–2002 5.0% 4,033 834 0.0% PEM PFO PSS

Liberty County! wetland loss

1.4% Loss in acres 113 1,673 1.2% 163 1.0% 0.8% 0.6%

Percentage Loss from 1992–2010 0.4% 0.2% 0.0% PEM PFO PSS

19 Montgomery County wetland loss

10.2% 261 10.0% Loss in acres 2,286 9.8%

9.6% 9.4%

9.2% 537 Loss from 1992–2010 Percentage 9.0% 8.8% 8.6% 8.4% PEM PFO PSS

Waller County wetland loss

2.0% 68

1.5%

1.0% Loss in acres Loss from 1992–2010 3 0.5% 7 0.0% PEM PFO PSS

20 Appendix C. Atlas of wetland loss

Index sheet

21 22 23 24 25 26 27 28 29 30 31 32 33 Appendix D: Cowardin Classification

Map codes of wetland habitat types used in this and dominance types. Modifiers for water regime, application follow the classification system inClas - water chemistry and soils are applied to classes, sification of Wetlands and Deepwater Habitats of the subclasses and dominance types. Special modifi- United States, 1979, by Lewis M. Cowardin, et al. ers describe wetlands and deepwater habitats that The code structure is hierarchical, progressing have been created or highly modified by people or from systems and subsystems, to classes, subclasses, beavers.

Wetlands and deepwater habitats classification

System Subsystem Class Subclass |- Rb=Rock Bottom 1=Bedrock | 2=Rubble | |- Ub=Unconsolidated Bottom 1=Cobble-Gravel | 2=Sand | 3=Mud | 4=Organic | |-- 1=Subtidal------|- Ab=Aquatic Bed 1=Algal | | 3=Rooted Vascular | | 5=Unknown Submergent | | | |- Rf=Reef 1=Coral | | 3=Worm | | | |- Ow=Open Water/Unknown Bottom (Used On Older Maps) M=Marine------| | | |- Ab=Aquatic Bed 1=Algal | | 3=Rooted Vascular | | 5=Unknown Submergent | | | |- Rf=Reef 1=Coral |-- 2=Intertidal--- | 3=Worm | |- Rs=Rocky Shore 1=Bedrock | 2=Rubble | |- Us=Unconsolidated Shore 1=Cobble-Gravel 2=Sand 3=Mud 4=Organic

34 System Subsystem Class Subclass |- Rb=Rock Bottom 1=Bedrock | 2=Rubble | |- Ub=Unconsolidated Bottom 1=Cobble-Gravel | 2=Sand | 3=Mud | 4=Organic | |-- 1=Subtidal------|- Ab=Aquatic Bed 1=Algal | | 3=Rooted Vascular | | 4=Floating Vascular | | 5=Unknown Submergent | | 6=Unknown Surface | | | |- Rf=Reef 2=Mollusc | | 3=Worm | | | |- Ow=Open Water/Unknown Bottom (Used On Older Maps) | | E=Estuarine-- | | |- Ab=Aquatic Bed 1=Algal | | 3=Rooted Vascular | | 4=Floating Vascular | | 5=Unknown Submergent | | 6=Unknown Surface | | | |- Rf=Reef 2=Mollusc | | 3=Worm | | | |- Sb=Streambed 3=Cobble-Gravel | | 4=Sand | | 5=Mud | | 6=Organic | | | |- Rs=Rocky Shore 1=Bedrock | | 2=Rubble | | |-- 2=Intertidal-- |- Us=Unconsolidated Shore 1=Cobble-Gravel | 2=Sand | 3=Mud | 4=Organic | |- Em=Emergent 1=Persistent | 2=Nonpersistent | |- Ss=Scrub-Shrub 1=Broad-Leaved Deciduous

35 System Subsystem Class Subclass | | 2=Needle-Leaved Deciduous | 3=Broad-Leaved Evergreen | 4=Needle-Leaved Evergreen | 5=Dead | 6=Indeterminate Deciduous | 7=Indeterminate Evergreen | |- Fo=Forested 1=Broad-Leaved Deciduous 2=Needle-Leaved Deciduous 3=Broad-Leaved Evergreen 4=Needle-Leaved Evergreen 5=Dead 6=Indeterminate Deciduous 7=Indeterminate Evergreen |- Rb=Rock Bottom 1=Bedrock | 2=Rubble | |- Ub=Unconsolidated Bottom 1=Cobble-Gravel | 2=Sand |--1=Tidal------| 3=Mud | | 4=Organic | | | |-*Sb=Streambed 1=Bedrock | | 2=Rubble | | 3=Cobble-Gravel |--2=Lower | 4=Sand | Perennial---- | 5=Mud | | 6=Organic | | 7=Vegetated | | | |- Ab=Aquatic Bed 1=Algal R=Riverine--- |--3=Upper | 2=Aquatic Moss | Perennial---- | 3=Rooted Vascular | | 4=Floating | | Vascular | | 5=Unknown | | Submergent |-4=Intermittent- | 6=Unknown Surface | | | |- Rs=Rocky Shore 1=Bedrock

|- Rb=Rock Bottom 1=Bedrock | 2=Rubble | |- Ub=Unconsolidated Bottom 1=Cobble-Gravel | 2=Sand | 3=Mud | 4=Organic | |-- 1=Limnetic----- |- Ab=Aquatic Bed 1=Algal | | 2=Aquatic Moss | | 3=Rooted Vascular | | 4=Floating Vascular | | 5=Unknown Submergent | | 6=Unknown Surface | | | |- Ow=Open Water/Unknown Bottom (Used On Older | Maps) L=Lacustrine----| | | |- Rb=Rock Bottom 1=Bedrock | | 2=Rubble | | | |- Ub=Unconsolidated Bottom 1=Cobble-Gravel | | 2=Sand | | 3=Mud | | 4=Organic | | | |- Ab=Aquatic Bed 1=Algal | | 2=Aquatic Moss

Modifiers |- A=Temporarily Flooded |- B=Saturated |- C=Seasonally Flooded |- D=Seasonally Flooded/Well Drained |- E=Seasonally Flooded/Saturated |- F=Semipermanently Flooded |--Non-Tidal------|- G=Intermittently Exposed | |- H=Permanently Flooded | |- J=Intermittently Flooded | |- K=Artificially Flooded | |- W=Intermittently Flooded/Temporary (Used On Older Maps) | |- Y=Saturated/Semipermanent/Seasonal (Used On Older Maps) | |- Z=Intermittently Exposed/Permanent (Used On Older Maps) Water Regime- | |- U=Unknown | | | | |- K=Artificially Flooded | |- L=Subtidal | |- M=Irregularly Exposed | |- N=Regularly Flooded |--Tidal------|- P=Irregularly Flooded

|- 1=Hyperhaline |- 2=Euhaline |--Coastal |- 3=Mixohaline (Brackish) | Salinity------|- 4-Polyhaline | |- 5=Mesohaline | |- 6=Oligohaline | |- 0=Fresh | | Water-Chemistry | | |- 7=Hypersaline |--Inland |- 8=Eusaline | Salinity------|- 9=Mixosaline | |- 0=Fresh | | | |--Ph Modifiers |- A=Acid For All |- T=Circumneutral Fresh Water------|- I=Alkaline

Soil------|- G=Organic |- N=Mineral

|- B=Beaver |- D=Partially Drained/Ditched Special Modifiers------|- F=Farmed |- H=Diked/Impounded |- R=Artificial Substrate |- S=Spoil |- X=Excavated U = Uplands

40 Appendix E: Freshwater wetland loss geo-spatial data processing

The analysis and mapping of wetland loss to ferent projections to allow users to correctly overlay development involves at its simplest level compar- vector data to raster imagery stored in different ing the 1992 National Wetland Inventory (NWI) coordinate systems. polygons with the most recent aerial photography The quarter quads were used as a layer to keep available (2010). Development has a markedly track of the reviewed and marked wetland polygons. different pattern than undisturbed wetlands, which makes it easy to delineate the developed area (see 1.3 Aerial photography Figs. 7 through 11 in the text). The 2010 NAIP 1M CIR aerial images from the To perform the geospatial processing, 2010 NAIP Texas Natural Resources Information System were color infrared (CIR) images were used as backdrop used for each county. These photos have the follow- imagery where 1992 NWI data in digital format ing projection and datum: were merged, overlaid, and edited using heads-up digitizing (on-screen). NWI polygon features were Universal Transverse Mercator projection, zone cut to reflect destruction of wetlands due to urban 15. Datum: NAD 1983. Units: meters. development or other causes. NWI attribute tables 1.4 Floodplain data were modified to include a field that tracks polygon change. 100-year floodplain data was taken from National Flood Hazard Layer dataset, which is a compilation Other fields were added to individual NWI of the effective Digital Flood Insurance Rate Map dataset attribute tables before merging, to facilitate database and Letters of Map Changes for Texas. analysis and exporting of detailed data at different levels: quarter quads, county, watershed, or the For Harris, Waller, and Liberty Counties, 2012 whole study area. The entire geoprocessing effort is floodplain data were used; for Galveston, Fort Bend, described below. Brazoria, Chambers, and Montgomery Counties, the 1996 FEMA Q3 floodplain data were used. A DVD of the 2012 data was ordered from FEMA’s 1. Input data product catalog: https://msc.fema.gov/webapp/wcs/ 1.1. Study area stores/servlet/StoreCatalogDisplay?storeId=10001&- catalogId=10001&langId=-1&userType=G The Greater Houston Metro Area is covered by 600 NAIP quarter quadrangles. The study area cov- 2. Geospatial processing ers the eight counties surrounding and including Harris County. ESRI ArcGIS Desktop 10 and 10.1 were used as main editing software to perform the entire 1.2 Wetlands vector data geo-processing. Wetland datasets from 1992 and 1993 were 2.1 NWI polygon editing downloaded in shapefile (.shp) format from the This step included modifying (cutting) polygon official NWI website (http://www.fws.gov/wetlands/ features where urban development or other change Data/Data-Download.html). was detected, based on aerial imagery (TNRIS 2010 All NWI datasets were merged using the same photos). The NWI shapefile was placed over the coordinate system, projection, and datum (Univer- NAIP imagery to look for wetlands that were filled sal Transverse Mercator (UTM) projection, Zone 15 by development. using NAD 83 datum, units: meters). Output vector Then, the attribute table of the NWI polygon layer datasets were re-projected and delivered using dif- was edited simultaneously to reflect the cause of

41 change. An additional field “DEV_2010” was added 3.2 Tabular output data to the attribute table, to input the appropriate cate- MS Excel was the main program used to produce gory of change (R, C/I, M etc.) tabular reports. The attribute table in GIS was also Each county NWI dataset was clipped out to used to perform statistics that were then incorpo- work on separately so as to simplify the database. rated into Excel. This process made the selection of any required attribute and the calculation of lost acreage and per- MS Excel centage much easier than with a combined dataset. The following are the categories of change: MS Excel files were first created by exporting the ARCGIS attribute table into a DBF file and then R: Residential reading and converting this file into an MS Excel C/I: Commercial/Industrial worksheet file format. Further calculations were F: Filled performed using Excel’s embedded mathematical W: Water functions. (See Appendix A for tabular results). D: Deforested The tables in Excel were sorted and calculated M: Mining according to the requirements (Fig. B2). For exam- Deforested values were not considered in the cal- ple, the exported data was sorted by attribute and culations of wetland loss but were marked for future then each of the attributes was summed to come reference and more investigation on the logging up with the value of each type. The number of each process in the marked areas. type of development was also calculated by sorting In addition, another field called “Calc_ the data by the development type (C/I, R, M, F, W). geo”/“Newacres” was created to calculate the The geometry of wetlands was calculated mostly geometry of the edited polygons in their respective in the NAD 1983 UTM zone 15N projected coor- coordinate systems, with the attributes of the acres dinate system. Wetland loss within floodplains was of wetlands lost. An image of the attribute table is calculated in NAD_1983_StatePlane_Texas_South_ shown in Figure B1. Central_FIPS_4204_Feet Projected Coordinate System.

Figure B1: Attribute table

3. Map output

3.1 Atlas of wetland loss Figure B2: Wetland polygon from the NWI overlain on 2010 color photo. The developed area is cut out and reclassified as A 24-page map book was created by defining “C/I” (Commercial/Industrial) in the attribute table. The unde- data-driven pages using ArcMap 10.1 and exporting veloped area is left blank in the new field for 2010 status. A the maps into high resolution jpeg format for the query method allows the change in the 1992–3 wetlands to purpose of the report. The grid was indexed and be calculated. custom sized to cover the study area.

42 Texas A&M AgriLife Extension Service AgriLifeExtension.tamu.edu

More Extension publications can be found at AgriLifeBookstore.org

Educational programs of the Texas A&M AgriLife Extension Service are open to all people without regard to race, color, sex, religion, national origin, age, disability, genetic information, or veteran status. The Texas A&M University System, U.S. Department of Agriculture, and the County Commissioners Courts of Texas Cooperating.

New